EP1978329A1 - Method for measuring the roundness of round profiles - Google Patents

Abstract

The method involves calibrating and determining a centre (Z-0) in the measuring field of the measuring device. The perpendiculars (r-1 to r-6, r-1' to r-6') from the centre to the tangents (T-1 to T-6, T-1' to T-6') and the distance from the centre to the tangents are calculated. The corner points of the polygon surrounding the spherical profile are calculated, and a contour is determined. A reference circuit is determined in relation to the contour such that the quadratic shape deviation of the contour reaches a minimum in relation to reference circuit. The reference circuit represents the smallest possible circuit that adapts to the contour, and represents the largest possible circuit that adapts to the contour. The reference circuit together with an additional circuit that is concentric to the reference circuit surrounds the contour at a minimal radial distance.

Description

The invention relates to a method for roundness measurement or measurement of the shape deviation of in its longitudinal direction advanced round profiles in rolling mills, in per se known manner using a measuring device with at least two laser scanners, each having a photosensitive sensor and a laser, at least three on the generated to survey the round profile fitting and the round profile enclosing, forming a polygon shadow edges and measured and calculated from the corresponding tangents.

In the steel industry so-called long products are rolled in specialized rolling mills in order to obtain the desired end products. If these long products are to be rolled into round bars, then they are usually rolled in a 3-roll finishing stand with several roll blocks (usually a 3-roll block) to the final dimension and run on a cooling bed for cooling. Usually, four roller blocks with three roller discs are used, wherein the center planes of the roller discs are pivoted from one to the next roller block by 60 °. Such bar steel often has polygonal shape deviations from the circular shape. The predominant form is "three-wave" or "six-wave".

If the diameter of such a polygonal bar steel is measured with a mechanical caliper or optically circumferentially, all the diameters determined can have the same value. In reality, the product is not round but rough / wavy. Such a product is called "equal thickness".

For the determination of so-called even thickness errors, calipers with support prisms were and are used. Depending on the waviness of the circumferential profile, different support angles are recommended.

The mechanical measurement of the roundness and their calculation have been explained and described for decades in relevant DIN standards, for example the German standards DIN ISO 4291 "Method for determining the roundness deviation", DIN ISO 6318 "Roundness measurement, terms and characteristics for the roundness" and DIN ISO 4292 "Procedure for measuring roundness deviations, two-point and three-point measuring methods".

The mechanical measurement of long products of the type in question with mechanical Lundheitsmessgeräten must be done offline. For this measurement, a sample must be clamped in a precision turning device. A feeler measures the radial deviations of the profile during the rotation. As a result, a graphic representation of the circumferential profile is obtained with the radii related to the respective angular degree. The evaluation of these peripheral profiles is described in detail in the aforementioned relevant standards.

In mechanical measurement in the measuring laboratory, an unlimited number of perimeter points can be determined by rotating the sample. On the other hand, in the production line where the product is transported longitudinally, the measurement of all the desired tangents must be made simultaneously to determine the profile of a local cross section. A mechanical online survey is therefore not possible.

An important output parameter in the evaluation and determination of out-of-roundness is the so-called reference circle and its center, to which all further steps of the measuring process refer. The standards describe four different types of determination.

In addition to the mechanical measuring devices, non-contact measuring devices have also been known for decades. For example, in the JP 56-117107 A a profile measuring method described in which the long product to be measured by means of laser beams is measured or scanned. Thus, this document describes, for example, that even with a uniform thickness a precise profile measurement is possible by applying first, second and third tangents to the outer circumference of an object whose profile is to be measured, and the difference between them Tangent specific circle and the profile of the long product to be measured whose profile measurement takes place. The application of the tangents is carried out with the help of laser beams or projection beams.

Also from the DE 100 23 172 A is a method for measuring the out-of-roundness of round products in question or round profiles known. In this case, a measuring device of three or more laser scanners, each having a photosensitive sensor and a laser used. The round product is illuminated by the laser beam from each of the laser scanners in such a way that the round product throws one or two shadow edges onto the associated sensor. For each of the shadow edges, a straight line running parallel to the laser beam is calculated. In addition, a circle is calculated from each of three determined degrees on which the straight lines lie as tangents. The calculation of the circle is repeated and the out-of-roundness is determined as the difference between the largest and smallest diameters of the circles.

This non-roundness determination has the disadvantage that with very small angle errors, the measured values are strongly falsified. This is especially the case if the tangents do not come to lie exactly on the maximum or minimum of the circumferential profile. Moreover, the centers of the circles in the room are not fixed, with the result that, for example, in the case of asymmetrical shape deviations, the profile determined has a periodic symmetry of the number of tangents and arrangements and does not reflect the true profile character.

The object of the present invention is to provide a method with which the profile and the runout can be determined as precisely as possible by means of a non-contact measuring device in the production line.

This problem is solved by the teaching of the claims.

In the method according to the invention, at least three shadow edges adjoining the round profile to be measured are generated by means of a measuring device equipped with at least two laser scanners. These laser scanners each have a photosensitive sensor and a laser.

Such a measuring device is known from the aforementioned JP 56-117107 A and also the DE 100 23 172 A known.

If only two laser scanners are available, then the round profile to be measured must lie completely within the light field of both laser scanners so that the required (at least three) shadow edges are generated and measured with these laser scanners.

If there are three laser scanners, it is sufficient if the round profile is only partially illuminated so that only one shadow edge is generated per laser scanner.

From these shadow edges, a tangential to the round profile fitting straight line or a tangent is calculated. The angles of the straight lines to each other are known. The laser scanners are arranged or selected such that at least three shadow edges a polygon is generated. The round profile to be measured is then within the area defined by this polygon. In the case of three shadow edges, this is a triangular area.

The method according to the invention is characterized, inter alia, by the fact that in step a) a center Z _{0 is} calibrated and fixed in the measuring field of the measuring device. The plane of the measuring field is expediently arranged perpendicular to the forward movement of the round profiles. The calibration of the measuring device and thus the determination of the center Z ₀ in the measuring field must be performed only once, for example, during assembly of the measuring device and / or after installation in the rolling mill before the online measurement is made of round profiles in question here. However, it is recommended to check the calibration from time to time and to repeat it if necessary.

In step b) of the method according to the invention, perpendiculars are determined from the center Z ₀ to the tangents measured during the method, and the distance between the center Z ₀ and the tangents is thereby calculated.

From the data calculated in step b), the vertices of the polygon enclosing the round profile are calculated in step c), whereby a contour is determined.

A reference circle is then placed in this contour in step d) of the method according to the invention. This reference circle can be defined in four different ways, namely:

i) The reference circle is placed so that the square shape deviation of the contour reaches a minimum to this reference circle.

ii) The reference circle is placed so that it represents the smallest possible circle that fits around the contour.

iii) The reference circle is placed so that it represents the largest possible circle that fits in the contour, or

iv) the reference circle is so placed with respect to the contour that this reference circle, together with another, concentric to the reference circle circle includes the contour with minimal difference in radius.

These alternative possibilities of defining the reference circle correspond to the definitions given in the aforementioned standards. In this regard, reference is made in particular to DIN ISO 6318 "Roundness Measurement". The definition of the reference circles is found there under 5. as follows:

• 5.1 circle of smallest deviation squares (LSC),
• 5.2 smallest circumscribed circle (MCC),
• 5.3 largest inscribed circle (MIC), and
• 5.4 Circles with the smallest ring zone (MZC).

After the calculation and definition of the reference circle, the diameter of the reference circle is calculated in the method according to the invention in step e). From the position in space, the reference center Zp is calculated, which represents the center point of the reference circle.

In step f) of the method according to the invention, at least two vectors from this reference center Z _p to the contour are calculated by this reference center. From this data the runout is determined.

Incidentally, not all generated shadow edges must be used to calculate the corresponding tangents. The selection of the shadow edges used to carry out the further calculation is made according to the requirement and the desired parameters, for example distance or angle of the tangents. The same applies to the number of verticals, which are used as the basis for further calculations.

It goes without saying, however, that the more tangents are measured at the same time, the more accurately the contour or the profile is imaged.

However, the number of laser scanners is usually limited for cost reasons and because of the limited size of the measuring device. In order to be able to display the profile as completely as possible, a contour simulation is preferably calculated from the available data on the polygon. Such a contour simulation can be expressed as a numerical approximation with polygons as a continuous function (Weierstrass approximation theorem). The contour simulation is preferably carried out with the aid of an adapted spline interpolation. Such smoothing calculations are otherwise well known to those skilled in the art.

This procedure makes it possible to subsequently apply all possible analyzes and measurement methods, taking into account the totality of the profile character. In particular, there are typical measured variables that are to be determined at a specific angle to the round profile or must be in a certain angular relationship to the measuring device. This is the case, for example, with a 3-roll stand in which the typical values GT and DT are very important in optimizing the setting of the individual stands, in particular for the last and second-last stand.

Another advantage of this contour simulation is the fact that it can be used for any number of laser scanners. The arrangement of these laser scanners and the angle division need not be even or regular, but can be selected as required. Important factors in this regard include, for example, the spatial conditions and the expected shape errors.

In the simplest case, only two vectors and thus the distance from Zp to the contour or to the contour simulation are calculated in the method according to the invention. Because the vectors are different Show directions and are in most cases of different sizes, this results already a value for the out-of-roundness. However, in most cases this is only an approximate value for out-of-roundness. Preferably, therefore, these vectors are determined and calculated such that they represent the minimum distance R _min and the maximum distance R _max of Zp to the contour or contour simulation.

Preferably, in step f) of the method according to the invention, one or more sets of 2, 3 or more vectors (V _GT1 , V _GT2 and V _GT3 or V _DT1 , V _DT2 and V _DT3 ) extending from Zp to Contour contour or contour simulation. The vectors of a sentence in particular include the same angle between them, e.g. B. 60 ° or 120 °. In the case of 120 °, such a set thus consists of three vectors.

Preferably, the vectors of one set point from the reference center point Zp towards the nip of the last stand, while the vectors of another set point from the reference center Zp towards the roll center of the last stand.

In the case of multiple sets of vectors, the angle included by the vectors is preferably the same for all sets of vectors. In addition, the vectors of one set are preferably rotated relative to the vectors of the other sets. For example, in particular, two sets of three vectors each can be calculated, with the vectors of each set including an angle of 120 °. If, for example, a vector of the first set of Zp points in the 0 ° direction (this reference direction must of course be specified) of the measuring field (more precisely the plane of the measuring field) up to the contour or contour simulation, then the two other vectors in 120 ° or 240 ° direction up to contour or contour simulation. The three vectors of the second set are twisted by 60 °, for example, so that from Zp in the 60 °, 180 ° and Show 300 ° direction. From this, the GT and DT values typical for 3-roll stands can be calculated without difficulty.

According to a further preferred embodiment, if due to the limited number of laser scanners (for example, only three laser scanners and only three shadow edges / tangents) not enough tangents can be calculated, the measuring device is rotated about the round profile, preferably in the form of an oscillating over 60 ° rotation ,

In this case the measurements are made at different times. In addition, the period of time which has elapsed between the individual measurements is determined. From this data, a motion vector for the round profile can be calculated so that it is possible to detect and compensate for the movement of the round profile. In other words, the measured data are computationally processed such that the measurements taken at intervals relate to the same reference center of the round profile.

It is also possible to choose the angular range of the rotational movement, depending on the number of shadow edges or tangents that can be generated, less than 60 °.

If, however, a sufficient number of laser scanners available and / or a sufficient number of shadow edges and thus tangents are generated, then the measuring device is preferably not rotated.

The data and measured values determined in carrying out the method according to the invention are transmitted in the usual way to an evaluation device and processed there. Such evaluation devices are known in nature, so that further explanations are unnecessary.

The invention will be explained in more detail below with reference to an example with reference to sketchy drawings which graphically represent the method steps according to the invention.

In the FIG. 1 is a cross section through a to be measured round profile shown, the outer contour is shown with a solid, bold line. A total of six shadow edges are applied to this round profile with the aid of three laser scanners, which give rise to the tangents T ₁ , T ₂ , T ₃ , T ₄ , T ₅ and T ₆ as well as T _{1 '} , T _2' , T _{3 '} , T _{4 '} , T ₅ , and T ₆ , lead. Tangent pairs T ₁ , T _{1 '} belong here; T ₂ , T _{2 '} ; T ₃ , T _{3 '} ; T ₄ , T _{4 '} ; T ₅ , T _{5 '} and T ₆ , T _6' each to a laser scanner. In total, six laser scanners are in use, whereby the round profile to be measured lies completely within the measuring field of these laser scanners.

Incidentally, the center Z _{0 of} the measuring field of the measuring device was more accurately determined and calibrated prior to the application of the shadow edges or tangents.

Even if a total of 12 tangents are applied to the round profile in the present example, this number of tangents T can be arbitrary. However, at least three tangents are required to form a polygon enclosing the round profile. The tangents have a known angular position to each other.

After the determination of the tangents, the perpendiculars r ₁ , r ₂ , r ₄ , r ₅ , r ₆ , r _{1 ',} r _2', r _{3 '} , r _4' , r _{5 '} and r _6' and thus the vertical Distance from Z ₀ to the respective tangents determined.

In the FIG. 2 It is shown that the tangents obtained in the manner described above form a polygon enclosing the round profile with the corners A to L. The polygon is in the FIG. 2 shown with a bold, solid line, while the round profile is represented by a dotted line.

In the FIG. 3 is shown, which form in the FIG. 2 shown polygon (in the FIG. 3 indicated by dotted line) Smoothing is obtained by means of a matched spline interpolation. This creates a contour simulation (solid bold line), which largely corresponds to the real round profile (solid thin line). In this way, usable data are obtained throughout the turn. In other words, data can also be determined for locations outside the values actually determined by the shadow edges or tangents.

In the FIG. 4 is shown graphically how a reference circle (dot-dash line) is placed in the contour simulation (solid bold line) such that the square shape deviations of the contour simulation to this reference circle (dot-dash line) reaches a minimum. For this reference circle whose diameter D _{ref is} calculated. The reference center Zp is determined from the position of the reference circle in the room.

In the FIG. 5 is shown as determined by this so determined center Zp two vectors, namely the minimum distance R _min and the maximum distance R _max from the reference center Zp for contour simulation and from these values, the out-of-roundness are determined. The determined extreme values can be anywhere on the contour simulation. They can thus also lie in angular positions, which are located in sections of the contour between the original measuring points.

In the FIG. 6 an alternative calculation method is illustrated graphically. The vectors V _GT1 , V _GT2 and V _GT3 on the one hand and V _DT1 , V _DT2 and V _DT3 on the other hand are calculated in step f) of the method according to the invention, extending from the reference center Zp in the direction of the roll disks of the last two 3-roll stands of a rolling mill It is assumed that the center plane of the roller disks of the penultimate rolling stand are at 0 °, 120 ° and 240 ° and that of the last rolling stand at 60 °, 180 ° and 300 °. The 0 ° or 180 ° level lies in the FIG. 6 in the plane which is perpendicular to the paper plane and through V _GT1 and V _DT3 , and is with the vertical dash-dotted line in the FIG. 6 indicated. In other words, the vectors V _GT1 , V _GT2 and V _GT3 point to the gap between the roll disks of the last mill stand, while the vectors V _DT1 , V _DT2 and V _DT3 point to the pressure point or to the roll center of the rolls of the last mill stand. Incidentally, this pressure point is located where the nip of the rolls of the penultimate roll stand is usually located.

From the vectors V _GT1 , V _GT2 and V _GT3 can be calculated by simple mathematical calculations for the setting of rolling stands of interest here in question value GT, which is a measure of length. The same applies to the vectors V _DT1 , V _DT2 and V _DT3 , from which the desired value DT, which also represents a length dimension, can be calculated.

These values are - as stated - authoritative for optimizing the setting of the individual 3-roll blocks and are rotated against each other over a fixed angle of 60 °.

Frequently, the measuring device required for the method according to the invention can not be arranged directly downstream of the last roll stand (for example for reasons of space) but only a certain distance downstream. There is now the problem that the finished rolled round profile is twisted on the path from the last roll stand to the measuring plane of the measuring device, wherein the angle by which the round profile rotates about its longitudinal axis on this route, is usually known in the individual rolling trains.

In the FIG. 7 is shown graphically how, despite the rotation of the round profile, the desired values GT and DT can be calculated. Because the twist angle that is in the FIG. 7 with α is known, the vectors described above are not in the direction of 0 °, 120 ° and 240 ° (applies to V _GT1 , V _GT2 and V _GT3 ) and 60 °, 180 ° and 300 ° (applies to V _DT1 , V _DT2 and V _DT3 ) from Zp to the contour simulation Definitely like that in the FIG. 6 is shown. Rather, these vectors are also rotated by the twist angle α. Thus, the vectors V _GT1 , V _GT2 and V _GT3 or V _DT1 , V _DT2 and V _DT3 rotated by Zp by the angle α are calculated from Zp to the contour simulation (bold solid line). The vectors V _GT1 , V _GT2 and V _GT3 are in the FIG. 7 each shown with an arrow with a solid line, while the vectors V _DT1 , V _DT2 and V _{DT3 are shown} with an arrow with a dashed line.

Thus, the typical values of GT and DT at the location of the last mill stand can also be determined by the same reference center Zp for arbitrary angular positions α taking into account all three vectors V _GT1 , V _GT2 and V _GT3 or V _DT1 , V _DT2 and V _DT3 . even if the measurement is performed only a certain distance after leaving the last mill stand.

Since each vector of the vectors V _GT1 , V _GT2 and V _GT3 or V _DT1 , V _DT2 and V _DT3 can be determined individually, the absolute delivery of the individual rolls of the respective roll stands can be determined with the method according to the invention. If, for example, the pressure point of a roller of a 3-roller block is offset further radially inwards than the pressure points of the two other rollers, then this can be determined according to the invention. It must then be corrected only the radial position of only one roller.

Claims (10)

A method for roundness measurement or measurement of the shape deviation of longitudinally advanced round profiles in rolling mills, in which per se known manner by means of a measuring device with at least two laser scanners, each having a photosensitive sensor and a laser, at least three voltage applied to the round profile to be measured and generate and measure the shadow profile enclosing the round profile, forming a polygon, and from this the corresponding tangents are calculated,
characterized in that a) before the measurement, a center (Z ₀ ) is calibrated and set in the measuring field of the measuring device, if such a center (Z ₀ ) has not yet been determined, b) vertical (r ₁ , r ₂ , r ₃ , r ₄ , r ₅ , r ₆ , r _{1 ',} r _2', r _{3 ',} r _4', r _{5 ',} r _6' ) from the center (Z _o ) to the tangents (T ₁ , T ₂ , T ₃ , T ₄ , T ₅ , T ₆ , T _{1 '} , T _2' , T _{3 '} , T _4' , T _{5 '} , T _6' ) determined and thus the distance from the center (Z _o ) to the tangents (T ₁ - T _{6 '} ) is calculated, c) the corner points (A to K) of the polygon surrounding the round profile are calculated, whereby a contour is determined, d) with respect to the contour of a reference circle is placed such that i) the square shape deviation of the contour to this reference circle reaches a minimum, ii) the reference circle represents the smallest possible circle that fits around the contour, iii) the reference circle represents the largest possible circle that fits in the contour, or iv) the reference circle, together with another circle concentric with the reference circle, encloses the contour with minimum radius difference, e) the diameter - (D _ref ) of the reference circle is calculated and from the position in space the reference center (Zp), which is the center of the Represents reference circle, is determined and f) at least two vectors from this reference center (Z _p ) are calculated to the contour and from this data the out-of-roundness is determined. Method according to claim 1,
characterized in that
which is followed by stage c) before stage d), stage c1), in which the polygon obtained in stage c) is expressed by a numerical approximation with polygons as a continuous function corresponding to the Weierstrass approximation theorem, whereby a smoothed contour simulation is obtained , and steps d) to f) are carried out using the contour simulation. Method according to claim 2,
characterized in that
the smoothing is carried out by means of an adapted spline interpolation. Method according to one of the preceding claims,
characterized in that
the vectors in step f) are calculated and determined such that it is the minimum distance (R _min ) and the maximum distance (R _max ) from the reference center (Z _p ) to the contour / contour simulation. Method according to one of claims 1 to 4,
characterized in that
determining one or more sets of 2, 3 or more vectors (V _GT1 , V _GT2 and V _GT3 or V _DT1 , V _DT2 and V _DT3 ) extending from the reference center (Z _p ) to the contour simulation; and the vectors of a sentence, in particular, enclose the same angle between them. Method according to claim 5,
characterized in that
in the case of multiple sets of vectors, the angle included by the vectors is the same for all sets of vectors, but the vectors of one set are rotated relative to the vectors of the other sets. Method according to claim 5 or 6,
characterized in that
show the vectors of one set from the reference center (Z _p ) in the direction of the nip of the last roll stand, while the vectors of another set point from the reference center (Z _p ) towards the roll center of the last roll stand. Method according to one of claims 5 to 7,
characterized in that
two sets of three vectors each are calculated, and the vectors of the first set are rotated relative to the vectors of the second set by a fixed angle of 60 ° to each other, and from this the values GT and DT typical for 3-roll stands are determined. Method according to one of the preceding claims,
characterized in that
the measuring device is rotated about the round profile and in particular performs a rotary movement oscillating over 60 °. Method according to one of claims 1 to 8,
characterized in that
the measuring device is not turned around the round profile.

Images